Alex Brandes ²

AI has crossed a threshold and knowledge work will never be the same. Last week we announced Gate AI, drop-in security and token savings for any agent. Here's why we built it: constellationnetwork.io/blog/why-we-bu…

Kimi k3 is really good. Wow.

A strategic blunder - how else can you describe it?

@alliekmiller Come find us at @_GateAI

Enterprises are in full-on cost management mode. They’re going to look to places like AWS, Fireworks, and OpenRouter to manage their costs down, while also prioritizing visibility into team usage, easy to set limits, auto cost management with smaller or OS models, speed to enabling state of the art models, zero data retention and other enterprise-friendly policies, access management, ease of use, and tool connections and control. If I were an orchestrator, I would go loud with marketing right about… now.

Some observations on Kimi: 1. It's a very good model! I don't think its performance can be explained away by distillation or anything like that. In agentic coding sessions, it seems pretty much on par with the best public models of Q1 2026. In my fairly limited use, it also seemed very token hungry. It's not obvious to me that this model is actually that cheap to run. 2. I am personally surprised the Chinese state continues to allow the open sourcing of models this good, given potential risks. To be clear, I *myself* might be fine with models presenting this level of marginal risk being open weight, but I am surprised that China is fine with it. I suspect the reason they are is 75% explained by strategic blindness/lack of AGI-pilledness (the CCP is very Yann Lecun-y in its views of AI). The other 25% or so is their lack of compute for customer inference (making China's open-weight strategy an unintended byproduct of US export controls) and the normal Chinese strategy of aggressive exports. For the companies, as opposed to the government, the decision to open source is partially ideological and partially because they are behind, and they know that very few people would pay for sub-frontier models from China. 3. Open-weight models are inherently decelerationist, and I'm continually surprised to see the so-called "accelerationists" so excited about open-weight models. I suspect the reason they are is that they know open-weight models are effectively ungovernable, and they simply like the overall cloak of ungovernability open-weight models create over the whole of AI. It's not a bad strategy; it reminds me of James Scott's recounting of the hill people in "the art of not being governed." Still, in the end, open-weight models deter further AI capex. 4. One probable outcome of an open-weight-model-dominant world is full AI communism, which is precisely what China proposes: rather than a market product, AI is a "public good" which will ultimately be provided by the state as a kind of "digital public infrastructure." This future strikes me as a dystopian hellscape, but I've never met an open-weight models advocate who doesn't ultimately concede this is where things end. You'd be surprised how many 'accelerationists' lobbied me, while I was in government, to support an eleven or twelve-figure federally funded data center so that startups could train models at a subsidy and then give them away for free. There was no other way for AI to progress, they said. Perhaps this is the logical end state of things. Nonetheless, I find myself surprised to see supposed accelerationists excited about such an outcome. I think many of them just don't know what they're doing. Many accelerationists do not view the creation and serving of frontier models as a legitimate business. 5. I would guess that the Trump Administration will at some point realize that their best strategy here would be to create large amounts of regulatory risk around the use of open-weight Chinese models. You don't need to "ban open source" (one of the dumber motifs of AI policy discussion). You just need to direct every agency to issue soft law that creates FUD. "A Federal Reserve Advisory Bulletin found that there may be backdoors in Chinese AI models." It needn't be that well justified. You just create enough regulatory risk that every regulated enterprise backs off. You probably don't want to create so much regulatory risk that you scare off the hyperscalers from serving Chinese models; this will just drive startups to sketchier providers. There's a happy middle ground here. I'd assume they will do some version of this. 6. It's probably true that open-weight models of this capability make the world a bit more dangerous, but not so much more that you'll really notice. At some point the models will be capable enough that you will notice. "A nonliving, invisible, dangerous, and infinitely self-replicating agent escaped from a Chinese lab," you say? Color me shocked.

Big if true

I put 48 million tokens through Claude in the last hour. You hit your limit before lunch. This is my own Gate AI dashboard, past week: 358.97M tokens across Claude, GPT-5.5, and Grok. 7,109 messages. Total spend: $57.23. One gateway, every model in one place, prompts compressed before they hit the meter, every call screened and fingerprinted on-chain. Running out of tokens is a single-provider problem. Bookmark this for the day your model cuts you off. And reply with your monthly AI bill: if it's over $57, I'll tell you what's eating it.

@ollama @jmorgan 85%?

Open models are already being used in the enterprise. Over 85% of the Fortune 500 companies already use Ollama to fulfill specific tasks. @jmorgan Why open models and Ollama? Ownership. Open models are yours to keep, customize, and optimize. Affordable. Run it the way you like, in your own environment. Private. Your data belongs to you.

@kevincodex Crypto survivors know - heads down, keep building 🫡

new motivational qoute unlocked: "but if there’s one thing i’ve learned from the last decade of building in this space, it’s that when things feel the worst, the best thing to do is just put your head down and build." will keep building much respect sir!

@thdxr Love your posts and apologies for shilling but this is exactly what we’re working on @_GateAI

prompt injection risk is so annoying man we could do so much cool automation if this wasn't a problem

@alexocheema While I 100% support local AI and am excited to see what Apple releases next - I have to say, this is some weird fear mongering.

If you don’t have at least 1TB VRAM and a VPN to torrent model weights from China, you are asleep at the wheel, and ignoring the uncomfortable truth that open weight models may be banned this year. Losing access to frontier models will be depressing and make a lot of people feel helpless - a few will have access to 100x productivity, and you will be stuck. Personally I cannot imagine doing work without frontier-level models right now. Think of the narrative: Chinese model (GLM-6?) with Fable capabilities, no guardrails. It’s plausible. Everyone is underpricing the chance that open models get banned. What are you going to do if it happens?

@RobertJBye Why?

The more you think about it, the cooler you realise this is.

A new American lab has entered the open-weight AI race with Inky, a nearly 1T parameter multimodal model. It doesn’t appear to be quite frontier-level on benchmarks yet, but it’s great to see another serious entrant pushing open models forward. I’m excited to try this one out on @_GateAI soon!

I’m not sure why I need this but I definitely need this.

A new Gate blog is out from @berg_dave42 Deploying AI everywhere makes work faster, not better - real efficiency comes from a single layer under every AI client that lets you secure it, see what it costs, and prove what it did.

Everyone gets Pro now. New sign-ups start with a 14-day Pro trial, and every existing free account has been upgraded to a Pro trial automatically. Nothing to do on your end. Go build something.

@VosDerek @_GateAI Sent you a DM.

@_GateAI @codebrandes Advanced setups have problems using your product. Lots of bugs. Who do we speak to?

The first post from our new blog is live: some of the vision behind Gate, from our CTO @codebrandes. AI's multi-model era is already here. You need an accountability layer. constellationgate.ai/blog/pick-your…

@Dagnum_PI @BenJorgensen Yo dawg I heard you like tokens

@BenJorgensen Actually crazy you can use Crypto to pay for AI usage

Used $DAG to pay for Gate credits. Zero fees on the checkout. Ran my first real prompt in OpenAI Codex through it right after. $0.07 used. The rails handled screening and spend compression in the background without me touching anything. Pick your models. Own your rails. This is what it looks like when you're actually using it instead of just reading the post. $DAG $AIAI

@jun_song Let’s go!

The new engine for MLX is in its final stages of development. Just ran GLM-5.2 on a single MacBook (116GB) hitting 41.8 tok/s with a 256k context window. Quality loss is only around ~4%, which puts it right at the 3-4bit quality level. The tech behind this uses a newly introduced layered architecture. When I first started, I was getting 10 tok/s with Kimi-K2.6 (128GB, 1024 context). Now it is fully at production level. Been grinding on this for months. Feels great to see it finally coming out soon.

@Dagnum_PI Nice! Now do Grok 4.5

AIAI’s Constellation Network Launches Gate AI: AI security gateway delivering prompt injection protection, token cost reduction, and verified auditability through Constellation Digital Evidence #AI #AIAI #AISecurity #AIAIholdings buff.ly/6si3Emi